overexpression of fos-related antigen-1 in head and neck squamous cell carcinoma

ORIGINAL ARTICLE

Overexpression of Fos-related antigen-1 in head and necksquamous cell carcinoma

Flavia R. R. Mangone*, M. Mitzi Brentani*†, Suely Nonogaki‡, Maria Dirlei F. S. Begnami§, Antonio

Hugo J. F. M. Campos§, Fernando Walder{, Marcos B. Carvalho{, Fernando A. Soares§, Humberto

Torloni§, Luiz P. Kowalski§ and Miriam H. H. Federico*†

*Faculdade de Medicina da Universidade de Sao Paulo (FMUSP), Departamento de Radiologia, Disciplina de Oncologia, Sao Paulo,

Brazil, †Laboratorio de Investigacao Medica, LIM 24, Hospital das Clınicas FMUSP, Sao Paulo, Brazil, ‡Instituto Adolfo Lutz, Sao Paulo,

Brazil, §Hospital do Cancer, Sao Paulo, Brazil, and ¶Servico de Cirurgia de Cabeca e Pescoco, Hospital Heliopolis, Sao Paulo, Brazil

Summary

The activating protein-1 (AP-1) family of transcription factors has been implicated in

the control of proliferation and differentiation of keratinocytes, but its role in malignant

transformation is not clear. The aim of this study is to assess the pattern of mRNA

expression of jun-fos AP-1 family members in 45 samples of head and neck squamous

cell carcinomas (HNSCC) and matched adjacent mucosa by means of Northern blot

analysis. Transcripts of all family members were identified, except for JunB that was

detected only by means of reverse transcription polymerase chain reaction. Neither

c-Fos nor JunD or FosB mRNA differed between tumours and normal tissues. We

observed a strong Fos-related antigen-1 (Fra-1) and Fra-2 expression, but only Fra-1

mRNA densitometric values were higher in tumour, compared to normal adjacent

mucosa (t-test, P5 0.006). A direct relationship between the positive expression of

Fra-1 mRNA, above tumour median, was associated with the presence of compromised

lymph nodes (Fischer exact test, P50.006). In addition, Fra-1 protein staining was

assessed in a collection of 180 tumours and 29 histologically normal samples adjacent

to tumours in a tissue array. Weak reactivity, restricted to the basal cell layer, was

detected in 79% of tumour adjacent normal tissues, opposed to the intense reactivity of

cancer tissues. In the subgroup of oral cancers, we have observed a shift in Fra-1

immunoreactivity, as long as the number of patients in each category, cytoplasmic or

nuclear/cytoplasmic staining, was analysed (Fischer exact test, P5 0.0005). Thus, Fra-1

gene induction and accumulation of Fra-1 protein may contribute to the neoplastic

phenotype in HNSCC.

Keywords

AP-1, Fra-1, HNSCC, Northern blotting, tissue array

Received for publication:

13 May 2004

Accepted for publication:

19 January 2005

Correspondence:

Maria Mitzi Brentani

Faculdade de Medicina da USP

Departamento de Radiologia

Disciplina de Oncologia

Av. Dr Arnaldo

455 4˚ andar

Sala 4112

01246-903 Sao Paulo

SP

Brazil

Tel.: +55 11 3082 6580

Fax: +55 11 3082 6580

E-mail: mbrentani@lim24.fm.usp.br

Int. J. Exp. Path. (2005), 86, 205–212

� 2005 Blackwell Publishing Ltd 205

Squamous cell carcinomas (HNSCC) are the most frequent head

and neck neoplasms, the sixth most frequent cancer worldwide

and the fifth in Brazil (Vokes et al. 1993; INCA 2002).

Several molecular alterations have been described that are

thought to play a role in HNSCC development and progres-

sion, involving genes important to cell proliferation and inva-

sion. As a result, many oncogene products, growth factors and

receptors linked to proliferative pathways are frequently over-

expressed in HNSCC (Schraml et al. 1999; Leethanakul et al.

2000; Takes et al. 2001; Serewko et al. 2002; Freier et al.

2003). The ultimate effect of this sustained activation of

growth factors and receptors may be the stimulation of nuclear

transcription factors, such as the ‘activating protein-1’ (AP-1).

The AP-1 transcription factor is a dimeric complex

composed of the members of the JUN, FOS, activating tran-

scription factor (ATF) and musculoaponeurotic fibrosarcoma

(MAF) protein families. The fos family consists of four gene

products (c-Fos, FosB, Fra-1 and Fra-2), whereas the jun

family is made up of three gene products (c-Jun, JunB and

JunD). As a consequence, the AP-1 complex can be formed by

a multitude of heterodimers and homodimers, depending on

the stimulus and duration. During carcinogenesis, various

dimer combinations may be required at various stages of

tumorigenesis (Eferl & Wagner 2003). The ultimate effect of

this imbalance may be the reprogramming of gene expression,

which may switch the selection of target genes to those

expressed during the invasive process (Crawford & Matrisian

1996; Ozanne et al. 2000). In fact, the overexpression of

proteolytic enzymes, such as urokinase-type plasminogen acti-

vator (uPA) and metalloproteinases, was already described in

HNSCC (Johansson & Kahari 2000; Alevizos et al. 2001; O-

Charoenrat et al. 2001; Pasini et al. 2001; Pacheco et al.

2002; Nagata et al. 2003).

In squamous carcinoma, previous published studies found a

relationship between AP-1 activity, cell transformation and

malignant progression (Yuspa 1998). Concerning head and

neck tumours, few studies had examined AP-1 expression

(Ondrey et al. 1999; Pacheco et al. 2002; Serewko et al.

2002) and none focused all members of jun–fos family. Our

own previously published data indicated that c-Jun mRNA

was differentially expressed in HNSCC, compared to that in

adjacent normal tissue, and correlated with the expression of

uPA mRNA, implicating in a possible role of these factors in

invasion (Pacheco et al. 2002).

The purpose of the present study is to investigate the expres-

sion pattern of the remaining jun–fos family members in a

group of HNSCC tumour fragments, collected prospectively.

Northern blot analysis of tumours was performed for each one

of the AP-1 members, and the changes in gene expression

pattern were analysed in the context of clinical pathological

parameters. Because Fos-related antigen-1 (Fra-1) was the

only member found consistently overexpressed in tumours,

compared to normal mucosa, we assessed the correspondent

protein immunohistochemically, in a collection of HNSCC

paraffin blocks, by using tissue array technique.

Patients and methods

Patients and tumour samples

In this study, we analysed by using Northern blotting 45

operable HNSCC patients admitted at Head and Neck

Surgery Service of Hospital Heliopolis, Sao Paulo, Brazil.

None of the patients had received any previous treatment.

Tumour staging was determined according to the fifth edi-

tion of the UICC TNM Classification of Malignant

Tumours. The histological grade, lymph node status and

tumour site and width were obtained from the surgical

pathological report. The general characteristics of the

patients have been described in Table 1. After surgery, the

surgical specimens of tumour and adjacent mucosa were

immediately frozen and were stored in liquid nitrogen. All

patients were advised of the procedures and gave informed

consent to participate in the study.

Table 1 Clinical–pathological characteristics of the 45 HNSCC

patients whose tumour tissues were studied by using Northern

blot analysis

Characteristic Number Percentage

Sex

Male 41 91.1

Female 4 8.9

Age

Median 55 –

Range 30–80 –

Tumour size

pT1 + T2 10 22.2

pT3 + T4 35 77.8

Lymph node status

pN0 23 51.1

pN+ 22 48.9

Tumour site

Oral cavity 20 44.5

Larynx 19 42.2

Oropharynx 06 13.3

Stage

III 23 51.1

IV 22 48.9

206 F. R. R. Mangone et al.

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212

CDNA synthesis and cloning

The complementary DNA (cDNA) sequences of each human

AP-1 mRNA member (c-Jun, JunB JunD, c-Fos, FosB, Fra-1

and Fra-2) were generated with reverse transcription polymer-

ase chain reaction (RT-PCR) by using specific primers

designed by ‘Primer3 Output’ software (http://www.genome.

wi.mit.edu/egi-bin/primer/primer3) and total RNA was

obtained from the MDA-MB-231 cell line. The forward and

reverse primer sequences used, respectively, were – c-Jun:

5´-AGG AGG AGC CTC AGA CAG TG-3´ and 5´-TGT TTA

AGC TGT GCC ACC TG-3´; JunB: 5´-ACT CTT TAG AGA

CTA AGT GCG-3´ and 5´-GAA ACA GAC TCG ATT CAT

A-3´; JunD: 5´-GAG TGT TCG ATT CTG CCC TA-3´ and

5´-TGT GGA CTC GTA GCA AAC AA-3´; c-Fos: 5´-ATC TGT

GCG TGA AAC ACA-3´ and 5´-TCC AGC ACC AGG TTA

ATT CC-3´; FosB: 5´-ACC CTT TTC TGA TCG TCT CG-3´

and 5´-CTG CTC ACA CTC ACA CTC G-3´; Fra-1: 5´GCC

TGT GCT TGA ACC TGA G-3´ and 5´-GCT GCT ACT CTT

GCG ATG A-3´; Fra-2: 5´-CCC AGT GTG CAA GAT TAG

CC-3´ and 5´-CCC AGT GTG CAA GAT TAG CC-3´; b-actin:

5´-CAC TCT TCC AGC CTT CCT TCC-3´ and 5´-CGG ACT

CGT CAT ACT CCT GCT T-3´. PCR was performed with

200 ng of cDNA and 0.25mM of each primer. The PCR profile

was: 5 min at 95 ˚C; 35 cycles: 1 min at 95 ˚C, 1 min at 54 ˚C

and 1 min at 72 ˚C; final extension at 72 ˚C for 10 min. The

purified PCR products were then subcloned into the PCR�

4-TOPOTA Cloning� kit (Invitrogen, Rockville, MD, USA)

and were sequenced. The 18S rRNA cDNA in pBR322 (1.9kb

Sal I/Eco RI fragment) was provided by Dr Arnhein, New

York City, NY, USA (Arnhein 1979).

Probes

Probes were radiolabelled by using a red-prime DNA labelling

system (Amersham Corp., Amersham, UK) with [a-32P]dCTP

(Amersham Corp.) and approximately 2 · 106 cpm/ml of

radiolabelled probes were used for hybridization.

RNA isolation and Northern blot analysis

Frozen tumour fragments were pulverized and total RNA was

isolated by using TRIZOL reagents (Invitrogen). Total RNA

(15 mg each) was electrophoresed on 1% agarose gel and was

transferred to a Hybond-N membrane (Amersham Corp.).

After overnight hybridization at 42 ˚C in hybridization solu-

tion (5· SSPE, 50% formamide, 5· dextran, 5% Denhardt’s,

1% SDS and 100 mg/ml of Salmon Sperm DNA), filters were

washed twice with 2· SSPE + 0.1% SDS solution at room

temperature and twice with 1· SSPE + 0.1% SDS, one at

52 ˚C and the other at room temperature, and then were

exposed to a Kodak XAR-5 X-ray film at _70 ˚C with an

intensifier screen. Band intensities were quantified by means

of densitometric scanning and data were expressed as the ratio

of the specific mRNA to 18S rRNA. A total of 37 patients

were analysed for the presence of c-Jun mRNA expression, 32

for JunD and FosB, 37 for c-Fos and 29 for Fra-1 and Fra-2.

Semi-quantitative RT-PCR

JunB mRNA expression was analysed in total RNA samples

obtained from tumour and adjacent normal tissue of 15

HNSCC patients. Five micrograms of total RNA was treated

with 1 U of DNAse–RNAse free (Promega, Madison, WI,

USA), 10 mM MgCl2 and 1 mM DTT in a final volume of

100ml of TE, pH 7.4, for 10 min at 37 ˚C. The reaction was

stopped by adding 25 ml of stop mix (100 mM EDTA, 3 M

sodium acetate and 2% SDS) and was followed by phenol–

chloroform extraction and ethanol precipitation. For the RT

reaction, total RNA was incubated with 5 mM random

hexamer (Amersham Corp.) at 70 ˚C for 10 min, and then

with 1· Super Script II buffer (Promega) for 60 min at 42 ˚C

20 mM dNTP, 5 mM DTT and 25 U Super Script II in a 20ml

final volume. PCR was performed as described above. Num-

ber of cycles was optimized for both JunB and b-actin primers

in order to ensure that amplification was in the linear range

and the results were semiquantitative. The established ampli-

fication cycles were 35 and 25 cycles for JunB and b-actin,

respectively. PCR products were visualized by means of

electrophoresis on a 1.5% agarose gel stained with ethidium

bromide and were quantified by means of densitometry (Image

Master – IMVDS software, Amersham).

Tissue array

A total of 180 cases of HNSCC were retrieved from the

archives of the AC Camargo – Cancer Hospital. All tissues

had been fixed in formalin and embedded in paraffin and

HE-stained slides from each specimen were submitted to

pathological review in order to reconfirm diagnosis and selec-

tion of the blocks. Samples included 87 oral cavity cases (39

pN0, 48 pN+), 50 larynx tumours (25 pN0, 25 pN+) and 43

pharynx carcinomas (20 pN0, 23 pN+). Twenty-nine samples

of normal mucosa were also included. For the construction of

the HNSCC tissue array, new sections were obtained from the

representative paraffin blocks, and all the H&E–stained slides

for these cases were reviewed. A slide with representative

condition was selected from each case and an area was circled

on the slide. The corresponding formalin-fixed, paraffin-

embedded blocks were retrieved, and the area corresponding

to the selected area on the slide was circled on the block. Using

Fra-1 in HNSCC 207

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212

a tissue microarrayer (Beecher Instruments, Silver Spring, MD,

USA), the area of interest in the donor paraffin block was

cored twice with a 0.6-mm diameter needle and was trans-

ferred to a recipient paraffin block. Sections of 4 mm were cut

from HNSCC tissue array blocks, deparaffinized, rehydrated

and submitted to immunohistochemistry with Fra-1 antibody

(Santa Cruz Biotechnology, Santa Cruz, CA, USA; dilution

1 : 100). Staining for Fra-1 was visualized by using a standard

3,3´ diamino benzidine procedure. Fra-1 expression by more

than 10% of cells was considered positive. Specificity of the

immunoreaction was verified upon omission of the primary

antibody. Epithelia of five patients who underwent surgery for

reasons other than malignancy were collected for comparison,

including three tongue samples (haemangioma), one rhino-

pharynx (lymphoid hyperplasia) and one larynx (cyst).

Statistical analysis

The associations of categorical variables between tumour and

adjacent mucosa were investigated by means of the paired t-test,

Fisher exact test and the Box plot analyses, performed by using

SPSS 10.0 software (SPSS Inc., Chicago, IL, USA). Differences

were considered statistically significant when P� 0.05.

Results

Tumour fragments and adjacent normal mucosa of 45 patients

with HNSCC were examined for the expression of c-Jun,

JunD, JunB, c-Fos, FosB, Fra-1 and Fra-2 mRNA. Representative

Northern signals from tumours and their respective non-malignant

areas have been showed in Figure 1a. c-Jun, JunD and c-Fos

transcripts were, respectively, identified as single autoradio-

FosB -

a

b

- 3.8 kb

- 2.7 kbc-Jun -

JunD -

18S -

Fra-1 -

18S -

T

T

JunB

β-actin

N N N NT T T

N T N T N T N

- 1.9 kb

- 7.0 kb

- 4.4 kb

- 2.4 kb

- 1.9 kb

- 2.2 kb

- 1.9 kb

Fra-2 -

18S -

c-Fos -

18S -

- 1.9 kb

- 3.3 kb

- 1.7 kb

- 1.9 kb

Figure 1 AP-1 mRNA expression in head and neck squamous cell carcinoma. (a) Representative Northern blot assay. Total RNA was

extracted, electrophoresed on agarose gel and transferred to a nylon membrane, as described in the section entitled ‘Patients and

methods’ in the text. These filters were hybridized with complementary cDNA probes labelled with [a-32P] dCTP (2·106 cpm/ml) and to

rRNA 18S (5· 105 cpm/ml). The AP-1 mRNA expression was estimated as the densitometric value relative to corresponding rRNA 18S signal.

(b) Representative JunB semiquantitative RT-PCR assay. N – adjacent normal mucosa; T – tumour samples. AP-1, activating protein-1.

208 F. R. R. Mangone et al.

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212

graphic bands of 2.7, 1.9 and 2.2 kb. Two Fra-1 transcripts

(3.3 and 1.7 kb) and three Fra-2 (7, 4.4 and 2.4 kb) were

detected. Because the Northern blot method was not sensi-

tive enough to quantify low levels of JunB mRNA, this

isoform was analysed by using RT-PCR. The analysis of

14

12

10*

*

**

**

***

8

6

AP-

1 m

RN

A e

xpre

ssio

n

4

2

0

n = 31JunD c-Fos FosB

TumourAdjacent mucosa

Fra-1 Fra-232 34 37 31 32 28 29 28 29

Figure2 Box plot analysis representing the mRNA expression of JunD,

c-Fos, FosB, Fra-1 and Fra-2 mRNA normalized to 18S rRNA in

HNSCC and matched adjacent mucosa. Boxes, 25th, 50th and 75th

percentiles; bars, 10th and 90th percentiles;˚, outlier values (1.5–3 box-

lengths from the 75th percentiles);*, extreme values (>3 box-lengths

from the 75th percentiles). Two patients with extreme values for Fra-1

expression, both for tumour and for mucosa, were excluded from the

graph for better visualization. HNSCC, head and neck squamous cell

carcinomas; Fra-1, Fos-related antigen-1.

3.0

2.5

2.0

1.5

1.0

n = 12

n = 5

Fisher exact test, P = 0.006

n = 9

n = 2

pN0 pN+

negFra-

1 m

RN

A e

xpre

ssio

n

Pos

0.5

0.0

Figure 3 Dispersion of patients according to Fra-1 mRNA expression

and lymph node status. Each point represents a patient and the spotted

line, the median values of tumour Fra-1 expression, above which,

tumours were considered Fra-1-positive (pos) and bellow, negative

(neg). The analysis of the distributionof values of each group showed a

close relationship between Fra-1 expression and the presence of lymph

node invasion (Fisher exact test, P50.006). Three patients with

extreme values were taken off the graphic for better visualization.

Fra-1, Fos-related antigen-1.

a

b

c

d

e

Figure 4 (a) Part of the tissue array slide containing 180 HNSCC

and 29 adjacent normal mucosa samples immunohistochemically

stained against Fra-1 protein (magnification: ·22.5). (b)

Surrounding normal tissue showing staining of cells in basal

layer (magnification: ·400). (c) Well-differentiated oral cavity

squamous cell carcinoma, stage pN0, displaying concomitant

nuclear and cytoplasmic immunostaining (magnification: ·400).

(d) Moderately differentiated squamous cell carcinoma of the oral

cavity, stage pN0, with nuclear and more intense cytoplasmic

immunostaining (magnification: ·400). (e) Moderately differ-

entiated squamous cell carcinoma of the oral cavity, stage pN+,

with exclusively cytoplasmic immunostaining (magnification:

·400). HNSCC, head and neck squamous cell carcinomas.

Fra-1 in HNSCC 209

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212

PCR products by means of agarose gel electrophoresis

showed the presence of JunB mRNA in normal and tumour

tissues. The semiquantitative RT-PCR suggested that the

expression of JunB was more intense in the morphologically

normal adjacent tissue, compared to tumours, but differ-

ences were not statistically significant (Figure 1b).

Quantification of Northern signals, after normalization

with rRNA 18S, showed that c-Fos, FosB and JunD mRNA

levels varied strongly among tumour specimens. FosB and

c-Fos mRNA levels ranged from 0 to 2.68 (mean5 0.746 0.71)

and from 0.01 to 3.05 (mean5 0.776 0.78), respectively.

Weak to moderate expression of these mRNA was found in

approximately 60% of the tumours, whereas strong expression,

considered as a densitometric value above one (>1) was

observed in 40%. Concerning JunD, 75% of tumour samples

presented strong or very strong expression and only 25%

exhibited moderate expression. JunD mRNA densitometric

values ranged from 0.55 to 5.0, with a mean value of

1.766 1.22. Fra-1 values ranged from 0.35 to 2.75, with a mean

value of 1.286 0.7, whereas Fra-2 mRNA values ranged from

0.66 to 6.51 (mean5 2.76 2.0), with 58 and 84% of the tumours

presenting very strong expression of Fra-1 and Fra-2, respectively.

The degree of variation of AP-1 family genes in normal and

neoplastic tissue has been illustrated in Figure 2. No differences in

c-Fos, JunD or FosB expression were found between tumours and

normal adjacent mucosa, whereas Fra-1 mRNA was consistently

overexpressed in tumours, compared to normal controls

(P5 0.006). Concerning the distribution of Fra-2 densitometric

values among non-neoplastic and neoplastic tissues, they were

marginally different, as showed by paired t-test (P5 0.07).

Our next step was to associate the expression of each AP-1

family member with site, age and histological grade and no

significant associations were found between AP-1 and those

parameters. Concerning TNM stage, we did not find any

differences between the expression of c-Fos, JunB, JunD or

Fra-2 in tumours with compromised lymph nodes (pN+),

compared to tumours without (pN0) lymph node involvement.

When tumours were classified as Fra-1-negative or Fra-1-positive

(equal/below or above the median tumour mRNA expres-

sion, respectively), there was a direct correlation between

Fra-1 expression and lymph node invasion (Fisher exact

test, P5 0.006, Figure 3). In relation to tumour size, only

JunD mRNA was significantly lower in T1/T2, compared to

that in T3/T4 tumours (0.846 0.44 vs 2.076 1.22, P5 0.003).

Because we had found Fra-1 mRNA overexpressed in head and

neck tumours, the corresponding protein was studied, initially in a

small subgroup of 10 patients, by means of immunohistochemis-

try. Owing to intense immunoreactivity, a tissue array was per-

formed in a collection of 180 tumours, 29 tumour adjacent normal

samples and five normal mucosa from non-cancer patients (Fig-

ure 4). Almost all tumour samples (95%) were strongly immunor-

eactive for Fra-1-recognizing antibody, whereas tumour adjacent

normal tissue showed a weak immunoreactivity restricted to basal

cell layer. The results provided in Table 2 showed that the sub-

cellulardistributionofFra-1, inaddition,differedamong tumours,

compared to normal tissues. Very few cases were completely

negative; almost half the cases (79/180 cases) showed nuclear/

cytoplasm double staining, only cytoplasmic staining (61/180) or

onlymembranous staining (36/180).Exclusively nuclear reactivity

was never found in tumours.

A subgroup analysis of oral cavity tumours showed a clear shift

inFra-1 immunoreactivity, from asimultaneousnuclear and cyto-

solic toanexclusivelycytoplasmiclocalization(Table 2),aslongas

the number of patients with each staining pattern was compared

between pN0 and pN+ groups (P5 0.0005, Fisher exact test).

A tendency towards this behaviour was also observed in pharynx

Table 2 Summary of immunohistochemical staining against Fra-1 results in HNSCC tumours (n5 180) classified by means of site and

lymph node status (compromised, pN+ or not, pN0) and morphologically normal epithelium (n529)

Fra-1 immunoreactivity (number of patients)

Primary

tumour site Lymph node Negative

Exclusively

nuclear

Exclusively

cytoplasmic

Nuclear/

cytoplasmic

Exclusively

membrane

Total

n

Oral cavity pN0 1 0 5* 29* 4 39

pN+ 1 0 21* 17* 9 48

Pharynx pN0 1 0 8 8 3 20

pN+ 1 0 13 4 5 23

Larynx pN0 0 0 7 11 7 25

pN+ 0 0 7 10 8 25

Tumour adjacent

mucosa 6 4 3 16 0 29

*Fisher exact test, P50.0005.

210 F. R. R. Mangone et al.

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212

tumours (P5 0.15), but was not seen in larynx tumours. The

progressive accumulation of Fra-1 protein and the changes occur-

ring in its subcellular localization have been illustrated inFigure 4.

None of the five normal mucosa, derived from patients without

malignancy, presented a positive reactivity with Fra-1 antibody.

Representative examples of Fra-1 protein immunostaining in nor-

mal adjacent tissue, well-differentiated pN0 tissue, moderately

differentiated pN0 samples and moderately differentiated pN+

tissue were displayed, showing a progressive accumulation of the

protein and, in addition, the change in its subcellular localization.

Discussion

We, in this study, report a systematic analysis of mRNA

expression of the members of the AP-1 family in HNSCC.

Even if c-Fos, FosB, JunB and JunD mRNA densitometric

values, displayed by neoplastic tissues, did not differ from

correspondent values expressed by normal, matched tumour

adjacent mucosa, Fra-1 mRNA were overexpressed in tumours.

This is in agreement with our results showing an increased immu-

noreactivity of Fra-1 protein determined by means of tissue array.

In parallel, an increased stability of the Fra-1 protein may occur

(Casalino et al. 2003; Vial & Marshall 2003). In addition, we had

already described that another member of this family, c-Jun, was

overexpressed in HNSCC (Pacheco et al. 2002).

Our observations, showing that not only Fra-1 mRNA but

also Fra-1 protein is increased in HNSCC, are in accordance

with previous studies in HNSCC lines (Ondrey et al. 1999;

Serewko et al. 2002). The overexpression of Fra-1 protein had

also been described in a variety of epithelial neoplasias, includ-

ing thyroid cancer (Chiappetta et al. 2000), esophageal squa-

mous carcinoma (Hu et al. 2001), endometrial and prostate

cancer (Bamberger et al. 2001; Zerbini et al. 2003). Data reported

by Zajchowski et al. (2001) demonstrated a tight association of

Fra-1 expression with highly invasive breast cancer cell lines.

There are several evidences, indicating that Fra-1 plays an

essential role during the epithelial cell oncogenesis. However, the

exact molecular implications and the role played by this marker in

malignancy remain unclear. One possibility is that its overexpres-

sion results in the transcription of progression-associated genes

and induction of epithelial mesenchymal transition (Bergers et al.

1995; Kustikova et al. 1998). In addition, Fra-1 is thought, in

partnership with c-Jun, to drive the expression of genes, such as

uPA, uPAr and PAI-1 (Anderson et al. 2002). Our own pre-

ceding published data on AP-1 had suggested an association

between the concomitant increase of c-Jun and uPA mRNA

(Pacheco et al. 2002).

The preferential cytoplasmic/nuclear concomitant localization,

seen in pN0 oral cancers, compared to that in pN+ tumours, is in

accordance with recently reported Fra-1 staining pattern described

in neoplastic thyroid diseases (Chiappetta et al. 2000). These data

are in disagreement with others describing Fra-1 as predominantly

nuclear in head and neck (Serewko et al. 2002) and in esophageal

cancer cells (Hu et al. 2001). Apart from the fact that we have

studied patient tumours and not cell lineages, the reason for these

differences is not clear. Our present data showed that Fra-1 is

accumulated in the cytoplasm of 44% pN+ oral cavity tumours,

possibly trapped and inactive, as described for the cytosolic form

of AP-1 family members in keratinocytes (Briata et al. 1993).

Therefore, Fra-1 probably could not be responsible for the increase

in uPA mRNA. This is, moreover, in agreement with our previous

data, showing that, at least in oral cancer, uPA mRNA was

decreased in pN+ tumours, compared to that in pN0 tumours

(Pasini et al. 2001). However, this conflicts with the conception

that pN+ tumours express high levels of proteinases (Nagata et al.

2003). Alternatively, the activation of specific signalling path-

ways in lymph node-positive tumours may result in an imbalance

among AP-1 complexes, favouring other AP-1 dimers implicated

with oral cancer progression.

Apart from the positive immunostaining of Fra-1 in some

tumour membranes, surrounding tissues never displayed a

similar localization, suggesting that a fraction of newly made

Fra-1 can be secreted. Additional determinations are required

in order to clarify the reason for this unexpected localization.

In our study, it must be stressed that the histologically

normal appearing mucosa, adjacent to the tumour tissue,

was also positive for Fra-1, which never occurred in tissues

of patients without neoplasia. These findings may reflect early

changes occurring before the appearance of histologically

recognizable malignant tissue in HNSCC.

In conclusion, we have found Fra-1-specific transcripts

increased in almost all tumours. These accumulation of Fra-1

mRNA and protein seems to be a widespread event in head

and neck carcinomas. However, the molecular mechanism

involved needs to be studied in detail.

Acknowledgements

Technical assistance of C. Nascimento and C.T. Braconi was

gratefully acknowledged. This study was supported by Fapesp

02/01738-9 and CNPq.

References

Alevizos I., Mahadevappa M., Zhang X. et al. (2001) Oral cancer

in vivo gene expression profiling assisted by laser capture

microdissection and microarray analysis. Oncogene 20,

6196–6204.

Andersen H., Mahmood S., Tkach V. et al. (2002) The ability of

Fos family members to produce phenotypic changes in

Fra-1 in HNSCC 211

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212

epithelioid cells is not directly linked to their transactivation

potentials. Oncogene 21, 4843–4848.

Arnhein N. (1979) Characterization of mouse ribosomal gene

fragments purified by molecular cloning. Gene 7, 83–96.

Bamberger A.M., Milde-Langosch K., Rossing E., Goemann C.,

Loning T. (2001) Expression pattern of the AP-1 family in

endometrial cancer: correlations with cell cycle regulators.

J. Cancer Res. Clin. Oncol. 127, 545–550.

Bergers G., Graninger P., Braselmann S., Wrighton C., Busslinger M.

(1995) Transcriptional activation of the fra-1 gene by AP-1 is

mediated by regulatory sequences in the first intron. Mol.

Cell Biol. 15, 3748–3758.

Briata P., D’Anna F., Franzi A.T., Gherzi R. (1993) AP-1 Activity

during normal human keratinocyte differentiation: evidence for

a cytosolic of AP-1/DNA binding. Exp. Cell Res. 204, 136–146.

Casalino L., De Cesare D., Verde P. (2003) Accumulation of

Fra-1 in ras-transformed cells depends on both transcriptional

autoregulation and MEK-dependent posttranslational stabil-

ization. Mol. Cell. Biol. 23, 4401–4415.

O-Charoenrat P., Rhys-Evans P.H., Eccles S.A. (2001) Expression

of matrix metalloproteinases and their inhibitors correlates with

invasion and metastasis in squamous cell carcinoma of the head

and neck. Arch. Otolaryngol. Head Neck Surg. 127, 813–820.

Chiappetta G., Tallini G., De Biasio M.C. et al. (2000) FRA-1

expression in hyperplastic and neoplastic thyroid diseases. Clin.

Cancer Res. 6, 4300–4306.

Crawford H.C. & Matrisian L.M. (1996) Mechanisms control-

ling the transcription of matrix metalloproteinase genes in

normal and neoplastic cells. Enzyme Protein 49, 20–37.

Eferl R. & Wagner E.F. (2003) AP-1: a double-edged sword in

tumorigenesis. Nat. Rev. Cancer 3, 859–868.

Freier K., Joos S., Flechtenmacher C. et al. (2003) Tissue

microarray analysis reveals site-specific prevalence of oncogene

amplifications in head and neck squamous cell carcinoma.

Cancer Res. 63, 1179–1182.

Hu Y.C., Lam K.Y., Law S., Wong J., Srivastava G. (2001)

Identification of differentially expressed genes in esophageal

squamous cell carcinoma (ESCC) by cDNA expression array:

over expression of Fra-1, Neogenin, Id-1, and CDC25B genes

in ESCC. Clin. Cancer Res. 7, 2213–2221.

Instituto Nacional do Cancer (INCA), Brazilian Health Ministry,

Estimativas da incidencia e mortalidade por cancer do Brasil,

Ministerio da Saude Brasil (2002).

Johansson N. & Kahari V.M. (2000) Matrix metalloproteinases

in squamous cell carcinoma. Histol. Histopathol. 15, 225–237.

Kustikova O., Kramerov D., Grigorian M. et al. (1998) Fra-1

induces morphological transformation and increases in vitro

invasiveness and motility of epithelioid adenocarcinoma cells.

Mol. Cell Biol. 18, 7095–7105.

Leethanakul C., Patel V., Gillespie J. et al. (2000) Distinct pattern

of expression of differentiation and growth-related genes in

squamous cell carcinomas of the head and neck revealed by the

use of laser capture microdissection and cDNA arrays.

Oncogene 19, 3220–3224.

Nagata M., Fujita H., Ida H. et al. (2003) Identification of

potential biomarkers of lymph node metastasis in oral

squamous cell carcinoma by cDNA microarray analysis. Int.

J. Cancer 106, 683–689.

Ondrey F.G., Dong G., Sunwoo J. et al. (1999) Constitutive

activation of transcription factors NF-(k) B, an NF-IL6 in

human head and neck squamous cell carcinoma cell lines that

express pro-inflammatory and pro-angiogenic cytokines. Mol.

Carcinog. 26, 119–129.

Ozanne B.W., McGarry L., Spence H.J. et al. (2000) Transcrip-

tional regulation of cell invasion: AP-1 regulation of

a multigenic invasion programme. Eur. J. Cancer 36,

1640–1648.

Pacheco M.M., Kowalski L.P., Nishimoto I.N., Brentani M.M.

(2002) Differential expression of c-jun and c-fos mRNAs

in squamous cell carcinoma of the head and neck: associations

with uPA, gelatinase B, and matrilysin mRNAs. Head Neck 24,

24–31.

Pasini F.S., Brentani M.M., Kowalski L.P., Federico M.H. (2001)

Transforming growth factor beta1, urokinase-type plasminogen

activator and plasminogen activator inhibitor-1 mRNA expres-

sion in head and neck squamous carcinoma and normal adjacent

mucosa. Head Neck 23, 725–732.

Schraml P., Kononen J., Bubendorf L. et al. (1999) Tissue

microarrays for gene amplification surveys in many different

tumor types. Clin. Cancer Res. 5, 1966–1975.

Serewko M.M., Popa C., Dahler A.L. et al. (2002) Alterations in

gene expression and activity during squamous cell carcinoma

development. Cancer Res. 62, 3759–3765.

Takes R.P., Baatenburg de Jong R.J., Wijffels K. et al. (2001)

Expression of genetic markers in lymph node metastases

compared with their primary tumours in head and neck cancer.

J. Pathol. 194, 298–302.

Vial E., Marshall C.J. (2003) Elevated ERK-MAP kinase activity

protects the FOS family member FRA-1 against proteasomal

degradation in colon carcinoma cells. J. cell. Sci. 116, 4957–4963.

Vokes E.E., Weichsebaum R.R., Lippman S.M., Ki long W.

(1993) Head and neck cancer. N Engl. J. Med. 328, 184–194.

Yuspa S.H. (1998) The pathogenesis of squamous cell cancer:

lessons learned from studies of skin carcinogenesis. J. Derma-

tol. Sci. 17, 1–7.

Zajchowski D.A., Bartholdi M.F., Gong Y. et al. (2001) Identifica-

tion of gene expression profiles that predict the aggressive

behavior of breast cancer cells. Cancer Res. 61, 5168–5178.

Zerbini L.F., Wang Y., Cho J.J., Libermann T.A. (2003)

Constitutive activation of nuclear factor kB p50/p65 and

Fra-1 and JunD is essential for deregulated interleukin 6

expression in prostate cancer. Cancer Res. 63, 2206–2215.

212 F. R. R. Mangone et al.

� 2005 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 86, 205–212